Search Results

The Oxygen Concentrator provide supplemental oxygen to patients who require supplemental O2, by separating nitrogen from room air, by way of a molecular sieve, it can be used in the home or health care facility. It is not intended to sustain or support life.

The Oxygen Concentrator is a portable concentrator which could utilize a molecular sieve and pressure swing adsorption to produce the oxygen. The device could separate nitrogen by absorbing through the molecular sieve when the room air enters the device, and allows the enriched oxygen to be collected. The device could provide approximately 90~96% of oxygen to patients on a continuous flow basis at a rate of 1L/min to 10L/min. The nasal cannula are not sold with the device, users need to purchase the accessories by themselves that are legally marketed. The products have not undergone biocompatibility assessment of condensate precipitates according to ISO18562-4, and it is not recommended to use humidification bottles. The maximum altitude the subject device can operate without degradation of concentration is 2000m.

I am sorry, but based on the provided text, I cannot extract the specific acceptance criteria and study details as requested. The document is a 510(k) summary for an Oxygen Concentrator, which primarily compares the subject device to a predicate device and outlines non-clinical tests performed. It does not contain sections detailing acceptance criteria for specific performance metrics with reported results, sample sizes for test sets, expert qualifications for ground truth establishment, adjudication methods, MRMC studies, standalone performance studies, types of ground truth used, or details about the training set.

The document states:

• "Clinical testing was not needed for this 510(k). The non-clinical performance testing described above is sufficient to support that the device can be used safely and effectively."

Therefore, there is no information about a study that would include elements like test set sample sizes, ground truth establishment, expert adjudication, or MRMC studies for this oxygen concentrator.

The "Performance Comparison" table in the document provides technical specifications and comparisons to the predicate device, but these are not presented as "acceptance criteria" with "reported device performance" in the format typically used for studies validating diagnostic or AI-based devices. The reported values are descriptive technical specifications rather than outcomes from a performance study against predefined acceptance criteria.

Therefore, I cannot provide the requested table and details because the provided text does not contain a study demonstrating device performance against explicitly defined acceptance criteria in the manner you described.

Ask a specific question about this device

The JUMAO Oxygen Concentrator is intended to provide supplemental oxygen to patients with respiratory disorders, by separating nitrogen from room air, by way of a molecular sieve. It is not intended to sustain or support life.

JMC5A Ni oxygen concentrator is made up of mainframe and flowmeter. It is an electronically operated device that separates oxygen from ambient air. It provides high concentration of oxygen directly to patient/user through a nasal cannula or other methods.

The Jumao Oxygen Concentrator is a Pressure Swing Adsorption (PSA) type oxygen concentrator which takes 120V ~ power source as power source. The output of oxygen is 0.5 to 5 liter per minute. Room air enters the piston type compressor via a series of filters for removing dust particles. The output compressed air is directed by a pneumatic valve into one of the two sieve beds which is full of adsorption material - molecular sieve. Nitrogen is adsorbed by the molecular sieve as the pressure increases; oxygen flows through the molecular sieve and concentrates at the sieve bed bottom. The enriched oxygen is divided into two streams; one stream enters a storage tank. The pressurized oxygen is regulated down to the suitable pressure, an adjustable flow meter and out to the patient. At the same time the second bed is in exhausted status, the molecular sieve desorbs nitrogen as the pressure decreases; another oxygen stream from first bed enters the bottom of the second bed, promotes purging the nitrogen and is exhausted into the atmosphere. Two sieve beds exchange the role of oxygen concentration and continue to produce 93% oxygen to the patient.

This document is a 510(k) premarket notification for an Oxygen Concentrator (Model: JMC5A Ni) and primarily focuses on demonstrating substantial equivalence to a predicate device. It is not a study report designed to prove a device meets specific acceptance criteria through clinical or large-scale non-clinical performance studies with detailed statistical analysis.

Therefore, many of the requested details about acceptance criteria, study methodologies, sample sizes for test/training sets, expert involvement, and ground truth establishment are not present in this regulatory submission document as they would be in a dedicated performance study report.

The document states, "No clinical testing has been performed." and primarily relies on non-clinical (bench) testing and a usability study to support substantial equivalence. The "acceptance criteria" discussed are more akin to meeting specific established standards and performance specifications for an oxygen concentrator rather than criteria for an AI/algorithm-driven device's performance metrics (like sensitivity, specificity).

Below, I will extract the available information, noting where the requested details are not provided due to the nature of the document.

Device Under Review: JUMAO Oxygen Concentrator (Model: JMC5A Ni)
Device Classification: Class II, Product Code CAW (Portable Oxygen Generator)
Indication for Use: "The JUMAO Oxygen Concentrator is intended to provide supplemental oxygen to patients with respiratory disorders, by separating nitrogen from room air, by way of a molecular sieve. It is not intended to sustain or support life."

1. Table of Acceptance Criteria and Reported Device Performance

The acceptance criteria are not explicitly listed in a "table" format with reported performance against them as one might find for an AI algorithm's metrics (e.g., sensitivity, specificity). Instead, the document discusses compliance with recognized standards and certain performance specifications.

2. Sample size used for the test set and the data provenance

• Test Set (for performance/engineering tests): The document doesn't specify a "sample size" in terms of number of devices for specific performance tests (e.g., number of concentrators tested for oxygen concentration). It states "All the bench test results are provided in Performance Test Report," implying that the necessary testing in accordance with the standards was performed on a representative sample of devices.
• Data Provenance: Not explicitly stated, but given the manufacturer is based in China, it's highly probable the non-clinical testing was conducted in China or by labs retained by the manufacturer. It's retrospective in the sense that the tests were completed before the 510(k) submission.

3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts

• This is not applicable as there is no "ground truth" to be established by experts in the context of human-read images or data for an AI/algorithm. The device is an oxygen concentrator, and its performance is measured against engineering specifications and industry standards.

4. Adjudication method (e.g. 2+1, 3+1, none) for the test set

• Not applicable for the type of device and testing performed. This typically applies to human reader studies or image interpretation.

5. If a multi-reader multi-case (MRMC) comparative effectiveness study was done, If so, what was the effect size of how much human readers improve with AI vs without AI assistance

• Not applicable. This is not an AI/algorithm-assisted diagnostic device. No clinical studies were performed.

6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done

• Not applicable. This is not an AI/algorithm device.

7. The type of ground truth used (expert consensus, pathology, outcomes data, etc.)

• The "ground truth" for this device's performance is defined by the technical specifications and requirements outlined in recognized national and international standards for oxygen concentrators (e.g., ISO 80601-2-69, various IEC and ISO biocompatibility and electrical safety standards). Performance data (e.g., oxygen concentration, flow rate, noise level) is measured directly from the device against these established benchmarks.

8. The sample size for the training set

• Not applicable. This is not an AI/machine learning device.

9. How the ground truth for the training set was established

• Not applicable. This is not an AI/machine learning device.

Ask a specific question about this device

The Oxygen Concentrator is intended to be used by patients with require supplemental oxygen. The device can be used in home, institution environments. The device is not intended to sustain or support life. The device is intended for use in adults.

Not Found

The provided document is a 510(k) clearance letter from the FDA for an Oxygen Concentrator. It describes the device, its intended use, and confirms its substantial equivalence to a predicate device. However, this document does not contain any information regarding acceptance criteria or a study proving the device meets acceptance criteria.

Therefore, I cannot provide the requested information, including:

1. A table of acceptance criteria and the reported device performance
2. Sample size used for the test set and data provenance
3. Number of experts used to establish ground truth and their qualifications
4. Adjudication method for the test set
5. MRMC comparative effectiveness study results
6. Standalone performance information
7. Type of ground truth used
8. Sample size for the training set
9. How ground truth for the training set was established

This type of information is typically found in the 510(k) summary or the full 510(k) submission, which is not included in the provided text. The letter only refers to the submission and states that the device has been found substantially equivalent.

Ask a specific question about this device

The Oxygen concentrator provide supplemental oxygen to patients with respiratory disorders, by separating nitrogen from room air, by way of a molecular sieve, it can be used in the home or health care facility. It is not intended to sustain or support life.

The oxygen concentrator adopts pressure swing adsorption principle, which can separate oxygen, nitrogen and other gas from the air, at constant temperature. As soon as power is connected, the air is taken in and compressed by oilless air compressor through filtering, then the compressed air goes through the cooler and it is cooled. After that, the air is taken into absorption tower by control valve and the oxygen can be separated in the absorption tower. At the same time, the high purity oxygen is collected into the oxygen tank, and it goes through the flow meter and humidifier, finally the oxygen that meets medical standards can be supplied. Oxygen is generated by pure physical method.

The provided text is a 510(k) summary for an Oxygen concentrator (K222920). It primarily focuses on demonstrating substantial equivalence to a predicate device through non-clinical testing and comparison of technical specifications. It does not describe acceptance criteria for an AI/ML powered medical device or a study involving such a device. The device in question is a physical oxygen concentrator, not an AI software.

Therefore, I cannot provide the requested information regarding acceptance criteria and a study to prove a device (specifically, an AI/ML medical device) meets those criteria, as the document does not pertain to such a device or methodology.

Ask a specific question about this device

This oxygen concentrator is intended for using as an oxygen supplement device in the professional healthcare facility and home healthcare environment. It provides high concentration of oxygen to persons requiring oxygen therapy. This device is to be used as an oxygen supplement and is NOT considered life-supporting or life-sustaining.

The Yuyue Oxygen Concentrator models 8F-3A/8F-5A is a portable concentrator which could utilize a molecular sieve and pressure swing adsorption to produce the oxygen. The device could separate nitrogen by absorbing through the molecular sieve when the room air enters the device, and allows the enriched oxygen to be collected. The subject device is intended use for adult population.

The subject device contains primary electrical components including main unit, compressor, fan, flowmeter and circuit breaker, the software is moderate level according to FDA software guideline.

This submission is for an Oxygen Concentrator, models 8F-3A and 8F-5A. The provided document does not contain information related to a study that proves the device meets acceptance criteria in the context of AI/ML performance, human-in-the-loop studies, or expert ground truth adjudication.

Instead, the document focuses on demonstrating substantial equivalence to a predicate device (L4 Oxygen Concentrator (EverFlo), K061261) through non-clinical performance data for regulatory clearance. This involves comparing technical specifications and performance against recognized standards, rather than evaluating AI performance with acceptance criteria typically found in machine learning contexts.

Therefore, many of the requested categories for acceptance criteria and study details cannot be filled from the provided text.

Here's an analysis based on the available information:

1. A table of acceptance criteria and the reported device performance

The document provides a comparison table for substantial equivalence, which includes various technical characteristics and a comparison to the predicate device. These characteristics serve as "acceptance criteria" for showing equivalence to the predicate, rather than explicit performance metrics against a defined threshold specific to a new AI/ML device.

2. Sample size used for the test set and the data provenance (e.g. country of origin of the data, retrospective or prospective)

Not applicable. The provided document details non-clinical laboratory testing and compliance with various standards, not performance results from a test set of patient data.

3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts (e.g. radiologist with 10 years of experience)

Not applicable. Ground truth establishment by experts is not relevant to this type of device and submission, which focuses on hardware performance and safety standards.

4. Adjudication method (e.g. 2+1, 3+1, none) for the test set

Not applicable. Adjudication methods are not relevant to this type of device and submission.

5. If a multi-reader multi-case (MRMC) comparative effectiveness study was done, If so, what was the effect size of how much human readers improve with AI vs without AI assistance

Not applicable. This device is an oxygen concentrator, not an AI/ML-driven diagnostic or interpretative tool.

6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done

Not applicable. This is a medical device, not an algorithm.

7. The type of ground truth used (expert consensus, pathology, outcomes data, etc)

Not applicable. Ground truth for AI/ML performance is not relevant here. The "ground truth" for this device's performance is its ability to meet the specified technical parameters (e.g., oxygen concentration, flow rate) and safety standards through physical testing and engineering measurements.

8. The sample size for the training set

Not applicable. This is not an AI/ML device; therefore, no training set is mentioned.

9. How the ground truth for the training set was established

Not applicable. This is not an AI/ML device.

Ask a specific question about this device

The Longfian Oxygen Concentrator, Model JAY-5AW is intended to be used by patients with respiratory disorders who require supplemental oxygen. A high concentration of supplemental oxygen is supplied and a nasal cannula is used to channel oxygen from the concentrator to the patient. The Longfian Oxygen Concentrator, Model JAY-5AW, can be used at home or healthcare environments. The Longfian Oxygen Concentrator, Model JAY-5AW, does not nor is it intended to sustain or support life. The device is intended for use in adults.

The Longfian Oxygen Concentrator, Model JAY-5AW adopts pressure swing adsorption principle, which can separate oxygen, nitrogen and other gas from the air, at constant temperature. As soon as power is connected, the air is taken in and compressed by oilless air compressor through filtering, then the compressed air goes through the cooler and it is cooled. After that, the air is taken into absorption tower by control valve and the oxygen can be separated in the absorption tower. At the same time, the high purity oxygen is collected into the oxygen tank, and it goes through the flow meter and humidifier, finally the oxygen that meets medical standards can be supplied. Oxygen is generated by pure physical method.

The device is capable of providing oxygen flow up to 5 LPM. The concentration of supplied oxygen is from 90%-96%. The weight of Jay-5AW is around 16kg (~351bs). There is no influence on indoor oxygen percent during the generator operating.

The biocompatibility contact classification of the device is as permanent contact externally communicating with dry and humified gas pathway.

The life-time/use life of the oxygen concentrator is set at 5 years or 20.000 hours after the product is sold, whichever comes first.

The reusable components:

• Intake air filter (two pieces, part number:GL-01)
• Secondary filter (one piece, part number:GL-02)
• nasal oxygen tube and
• humidifier.

JAY-5AW medical oxygen concentrator contains the following items:

• One (1) JAY-5AW Oxygen Concentrator Unit
• One (1) AC Power Supply
• One (1) Nasal Oxygen Tube
• One (1) Firesafe Cannula Valve
• One (1) Humidifier
• One (1) Operation Manual

The provided document is a 510(k) premarket notification for an Oxygen Concentrator, Model JAY-5AW. It details the device's technical specifications, indications for use, and a comparison to a predicate device. However, it does not contain information about a study proving the device meets acceptance criteria related to AI or a "human-in-the-loop" performance study of the type usually associated with the questions asked.

The document discusses "acceptance criteria" and "device performance" in the context of non-clinical bench testing for safety and performance (e.g., biocompatibility, electrical safety, EMC, software verification, and oxygen concentration/flow performance), not for an AI-driven diagnostic or assistive device that would involve human readers, ground truth establishment by experts, or MRMC studies.

Therefore, I cannot populate the table or answer the specific questions related to AI-driven device performance, sample sizes for test sets, expert adjudication, MRMC studies, standalone algorithm performance, or ground truth establishment in that context.

The document focuses on demonstrating substantial equivalence to a predicate device through engineering and safety testing of a physical medical device (an oxygen concentrator), not a software-based AI system.

Here's what can be extracted regarding the "acceptance criteria" and "device performance" as presented in this document for the Oxygen Concentrator:

Acceptance Criteria and Reported Device Performance (Based on the provided document for an Oxygen Concentrator):

Responses to Specific Questions (based on the provided document and its scope):

1. A table of acceptance criteria and the reported device performance: See table above.

2. Sample sizes used for the test set and the data provenance (e.g., country of origin of the data, retrospective or prospective):

   • The document describes bench testing of a physical device (oxygen concentrator prototypes/production samples), not a study involving patient data or AI algorithm validation with a "test set" in the common sense of AI/diagnostic imaging.
   • Therefore, no information on sample sizes of patient data or data provenance (country, retrospective/prospective) is provided or relevant to the type of testing detailed in this document. The testing describes electrical, functional, and biocompatibility tests performed on the physical device itself.

3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts (e.g., radiologist with 10 years of experience):

   • Not applicable. This document is for an oxygen concentrator, not an AI diagnostic device. There is no concept of "ground truth" derived from expert interpretation of medical images or data from a patient population, nor a need for radiologists or similar experts in this context. The "truth" for this device's performance is determined by meeting engineering and safety standards.

4. Adjudication method (e.g., 2+1, 3+1, none) for the test set:

   • Not applicable. There is no "adjudication method" involved as there are no expert interpretations to be reconciled.

5. If a multi-reader multi-case (MRMC) comparative effectiveness study was done, If so, what was the effect size of how much human readers improve with AI vs without AI assistance:

   • Not applicable. No MRMC study was performed as this is not an AI-assisted diagnostic device.

6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:

   • Not applicable. This is a physical medical device (oxygen concentrator), not an algorithm. The software mentioned (software verification and validation) pertains to the embedded control software of the device itself, not a standalone diagnostic algorithm.

7. The type of ground truth used (expert consensus, pathology, outcomes data, etc):

   • The "ground truth" for this device's performance is adherence to established international and national standards (e.g., IEC 60601 series, ISO 80601 series, ISO 18562 series, ISO 10993 series) and its stated design specifications (e.g., oxygen concentration, flow rate). These are technical performance metrics, not clinical diagnostic "ground truth."

8. The sample size for the training set:

   • Not applicable. There is no "training set" as this is not a machine learning/AI model.

9. How the ground truth for the training set was established:

   • Not applicable. No training set exists for this type of device submission.

Ask a specific question about this device

The OXYGEN CONCENTRATOR is intended to provide supplemental oxygen in a home, institutional, or travel environment.

Not Found

I am sorry, but the provided text is a 510(k) clearance letter for an Oxygen Concentrator. It does not contain any information about acceptance criteria, device performance, study details, human reader studies, or ground truth establishment relevant to an AI/ML device. Therefore, I cannot fulfill your request based on the given input.

Ask a specific question about this device

The Oxygen Concentrator is intended for the administration of supplemental oxygen. This device is not intended for life supporting nor does it provide any patient monitoring capabilities. The system will be operated by trained personnel, at home, in community health care and medical institutions.

The Oxygen Concentrator adopts the principle of pressure swing adsorption technology. At normal temperature, the device can continuously supply a high concentration of supplemental oxygen (93%±3%). The main function of the proposed device includes supplying supplemental oxygen. The device also includes alarms for high temperature, low oxygen output, and interruption of power. The proposed device includes three models, which are M50, M40 and M30. The only difference of the three models is flow rate.

The provided document pertains to a 510(k) premarket notification for an Oxygen Concentrator. As such, it describes a medical device, not an AI/ML algorithm. Therefore, many of the requested details regarding acceptance criteria, study design for AI performance, expert adjudication, and ground truth establishment are not applicable to this document.

The document focuses on demonstrating substantial equivalence to a predicate device through non-clinical testing and comparison of technical characteristics.

Here's a breakdown of the relevant information from the document, addressing the applicable parts of your request:

1. Table of acceptance criteria and the reported device performance

The document does not present "acceptance criteria" in the typical sense of a target performance metric for an AI algorithm. Instead, it describes compliance with established medical device standards and comparison to a predicate device. The "performance" is demonstrated by meeting these standards and showing similar technical characteristics to the predicate.

Below is a table summarizing the technical characteristics of the proposed device and how they compare to the predicate, which serves as the basis for the substantial equivalence determination. The "acceptance criteria" here are implied by the predicate device's characteristics and relevant standards.

Note on "Acceptance Criteria" for a non-AI/ML device: For this type of device, "acceptance criteria" are typically defined by compliance with recognized standards (e.g., ISO, IEC, AAMI) and demonstration of substantial equivalence to a legally marketed predicate device. The document explicitly states: "Non clinical tests were conducted to verify that the proposed device met all design specifications as was Substantially Equivalent (SE) to the predicate device."

2. Sample size used for the test set and the data provenance

This is not applicable as there is no "test set" in the context of an AI/ML algorithm's performance evaluation. The "testing" refers to non-clinical bench testing to demonstrate compliance with standards and safety/performance specifications of the physical device.

3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts

This is not applicable. No expert-established ground truth for an AI/ML algorithm's performance is involved in this 510(k) submission.

4. Adjudication method (e.g. 2+1, 3+1, none) for the test set

This is not applicable. No expert adjudication for an AI/ML algorithm's performance is involved.

5. If a multi-reader multi-case (MRMC) comparative effectiveness study was done, If so, what was the effect size of how much human readers improve with AI vs without AI assistance

This is not applicable. This submission is for a physical medical device (oxygen concentrator), not an AI-assisted diagnostic or therapeutic tool.

6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done

This is not applicable. There is no algorithm to evaluate in a standalone manner.

7. The type of ground truth used (expert consensus, pathology, outcomes data, etc)

This is not applicable. For this type of device, "ground truth" relates to the physical and functional parameters of the device (e.g., oxygen purity, flow rate, pressure) as measured by calibrated equipment and verified against engineering specifications and industry standards. It's not a diagnostic or clinical "ground truth."

8. The sample size for the training set

This is not applicable. There is no AI/ML algorithm involved, and thus no training set.

9. How the ground truth for the training set was established

This is not applicable. There is no AI/ML algorithm involved, and thus no training set or associated ground truth.

Summary of the Study Proving Device Meets Acceptance Criteria (and Substantial Equivalence):

The "study" refers to non-clinical testing and comparative analysis rather than clinical trials or AI performance evaluations.

• Type of Testing: Non-clinical tests were conducted. These typically include bench testing for electrical safety, electromagnetic compatibility, oxygen concentration accuracy, flow rate stability, alarm functions, sound levels, and biocompatibility of patient-contacting materials.
• Standards Met: The device demonstrated compliance with several international and US standards, including:
  • AAMI / ANSI ES60601-1 (General requirements for basic safety and essential performance of medical electrical equipment)
  • IEC 60601-1-2 (Electromagnetic Compatibility)
  • ISO 80601-2-69 (Particular requirements for basic safety and essential performance of oxygen concentrator equipment)
  • ISO 10993-5 (Biological Evaluation - In Vitro Cytotoxicity)
  • ISO 10993-10 (Biological Evaluation - Irritation and Skin Sensitization)
  • VOCs and PM2.5 test (per EPA TO 15 and IP-10 method)
• Comparison to Predicate Device (K150930): A detailed comparison of technical characteristics (Table 1) was performed against the predicate oxygen generator.
  • Similarities were highlighted in classification, product code, regulation number, intended use principle, and oxygen purity.
  • Differences in flow rate range, sound level, oxygen purity warning threshold, electrical requirements, and outlet pressure were acknowledged and discussed. The submission argues that these differences do not raise different questions of safety and effectiveness because:
    • The proposed device's flow rate is within the predicate's range.
    • Even with a lower oxygen purity warning level (<82% vs <85%), both comply with ISO 80601-2-69.
    • The outlet pressure, while different, also complies with ISO 80601-2-69.
    • Electrical requirements, though different, comply with the relevant electrical safety and EMC standards.
• Conclusion: Based on the non-clinical tests and the comparison, the manufacturer concluded that the proposed device models are substantially equivalent to the predicate device.
• Clinical Study: The document explicitly states: "No clinical study is included in this submission." This indicates that the substantial equivalence determination for this device did not require human clinical data.

Ask a specific question about this device

Porous Media oxygen concentrator filters are replacement filters intended for use in oxygen concentrator machines to help remove contaminants, including air borne bacteria and other particulate debris from an air stream. When used with oxygen concentrator machines, the replacement filters may be used in the home, nursing home, patient care facility, etc.

Room air is drawn into the compressor through the bacterial intake filter. From the compressor, the air passes through the compressor filter, if one is installed on the machine, and proceeds to the sieve beds. The sieve beds condition the air, by removing nitrogen from the air stream, which results in a higher concentration of oxygen. The air then passes through the final filter before being supplied to the patient.

This 510(k) summary describes Replacement Oxygen Concentrator Filters manufactured by Porous Media. The submission focuses on demonstrating substantial equivalence to existing predicate devices.

Here's an analysis of the acceptance criteria and study information provided, framed around a typical AI/medical device study structure, even though this submission is for physical filters and not an AI device. I will address each point from your request using the provided text.

1. A table of acceptance criteria and the reported device performance

For these oxygen concentrator filters, "acceptance criteria" are primarily based on demonstrating identical technical specifications and intended use to legally marketed predicate devices. The "reported device performance" is the direct assertion of these identical specifications. The submission is a comparison study, not a performance study against predefined thresholds for a novel device.

The study essentially proves the device meets the "acceptance criteria" by showing a feature-by-feature match with the predicate devices, thereby establishing substantial equivalence, which is the primary regulatory pathway for 510(k) devices. There is no independent performance study detailed here, as the performance is assumed to be equivalent to the predicate.

The following points are typically relevant for AI/algorithm-based medical devices. Since this submission is for physical filters, these specific questions regarding AI algorithms and ground truth methods are not directly applicable to the provided text. I will address them by stating their non-applicability or reinterpreting them in the context of a physical device.

2. Sample size used for the test set and the data provenance (e.g., country of origin of the data, retrospective or prospective)

• Not Applicable: This submission is for physical filters. There is no "test set" in the context of an AI algorithm evaluation. The "data" here refers to the specifications of the predicate devices and the filters themselves. The comparison is against published and known specifications of existing, approved devices. No specific "sample size" of devices was tested for this submission beyond the comparison to the existing designs. The data provenance is essentially from the design and manufacturing specifications of both the predicate devices and the Porous Media filters. No geographical or retrospective/prospective distinction is relevant here.

3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts (e.g. radiologist with 10 years of experience)

• Not Applicable: There is no "ground truth" in the AI sense for these physical filters. The "ground truth" for the comparison is the published, legally established specifications of the predicate oxygen concentrator filters, which are accepted by the FDA. The experts involved would likely be engineers, quality control specialists, and regulatory affairs personnel from Porous Media and potentially consultants verifying the comparison, but not "experts establishing ground truth" in a diagnostic sense.

4. Adjudication method (e.g. 2+1, 3+1, none) for the test set

• Not Applicable: No adjudication method is described or required for comparing physical device specifications. The comparison is a direct, objective matching of technical parameters.

5. If a multi reader multi case (MRMC) comparative effectiveness study was done, If so, what was the effect size of how much human readers improve with AI vs without AI assistance

• Not Applicable: This is a submission for physical filters, not an AI-assisted diagnostic device. Therefore, no MRMC study, human readers, or AI assistance is involved.

6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done

• Not Applicable: This is for physical filters, not an independent algorithm.

7. The type of ground truth used (expert consensus, pathology, outcomes data, etc)

• Not Applicable (in the AI sense): For this submission, the "ground truth" is the established technical specifications and performance characteristics of the legally marketed predicate devices, as approved by the FDA in their original 510(k) clearances. This is a regulatory "ground truth" rather than a clinical one.

8. The sample size for the training set

• Not Applicable: There is no "training set" as this is not an AI device.

9. How the ground truth for the training set was established

• Not Applicable: As there is no AI training set, there is no ground truth establishment for it.

Ask a specific question about this device